feijoa.plugins package#

Submodules#

feijoa.plugins.plugin module#

Base plugin’s class module.

class feijoa.plugins.plugin.Plugin(subscribers=None)#

Bases: feijoa.utils.mixins.Observer, feijoa.utils.mixins.Subject

Base plugin’s class.

The idea of plugins is to optionally add new functionality to search oracles. The plugin inherits from the Subject and Observer classes. This means that he can be both a subject (receive notification from oracle), as well as an observer (send alerts to oracle).

Parameters: subscribers (list) – Initial array of subscribers.
Raises: AnyError – If anything bad happens.

property aliases#

property anchor#

attach(observer)#

Attach an observer to the subject.

Parameters: observer (feijoa.utils.mixins.Observer) –
Return type: None

detach(observer)#

Detach an observer from the subject.

Parameters: observer (feijoa.utils.mixins.Observer) –
Return type: None

notify(event, *args, **kwargs)#

Notify all observers about an event.

Return type: None

abstract update(event, subject, *args, **kwargs)#

Receive update from subject.

Args:

event:
Type of event.

subject:
Subject of notification.

Returns:
None

Raises:
AnyError: If anything bad happens.

feijoa.plugins.reducer module#

Reducer class module.

class feijoa.plugins.reducer.DomainReducer(gamma_osc=0.7, gamma_pan=1.2, eta=0.9, min_window_value=0.0, subscribers=None)#

Bases: feijoa.plugins.plugin.Plugin

A sequential domain reduction transformer bassed on the work by Stander, N. and Craig, K: “On the robustness of a simple domain reduction scheme for simulation‐based optimization”

Link: http://www.truegrid.com/srsm_revised.pdf

Based on: https://github.com/fmfn/BayesianOptimization

Parameters

gamma_osc (float) – Shrinkage parameter for oscillation. Typically, [0.5-0.7].
gamma_pan (float) – Panning parameter. Typically, 1.0.
eta (float) – Zoom parameter.
min_window_value (float) –

Raises

AnyError – If anything bad happens.

aliases = ('domain_reducer',)#

anchor = 'reducer'#